/* mi-analysis.do --- 
* 
* Filename: mi-analysis.do
* Description: Run some of the analysis code with multiple imputation tagged to it
* Author: Kai Arzheimer
* Created: Tue Apr 25 16:16:03 2017 (+0200)
* Version: 1.0
* Last-Updated: Mon Mar 26 10:58:47 2012 (+0200)
*           By:
*     Update #: 7
* reads: 
* writes: 
* Compatibility: 14
* 
*/

/* Commentary: 
* 
* 
* 
*/

/* Change log:
* 
* 
*/

/* Code: */


version 15
clear all

* Load combined dataset
use /home/kai/Work/afd/transformation/data/full-series-matched-mi , clear

* go to flong. Argh

mi convert flong 

* Include positions on market vs state and immigration. Multiple imputation is attractive because more waves (and cases, obs) can be used

mi estimate , saving(mimymodelstwoadvanced-days, replace) esample(inmisample): meqrlogit afdvote c.zintdate i.male i.job i.educ i.region i.religion i.unemployed i.agecat c.t24  c.t114 c.t154 c.zintdate#i.male c.zintdate#i.job c.zintdate#i.educ  c.zintdate#i.region c.zintdate#i.religion c.zintdate#i.unemployed  c.zintdate#i.agecat c.zintdate#c.t24 c.t114#c.zintdate c.t154#c.zintdate || intdate:

* save /home/kai/Work/afd/transformation/data/full-series-matched-mi-marked , replace 

scons_returncode "`1'"





/*  mi-analysis.do ends here */
